INDEX

Explanations

measurement-related terms

oai_token-act-pair · gpt-3.5-turbo

terms related to measurement and quantification

oai_token-act-pair · gpt-4o-mini Triggered by @bot

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 4-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.4.hook_resid_pre

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

Skylion007/openwebtext

Features

24,576

Data Type

torch.float32

Hook Point

blocks.4.hook_resid_pre

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Hook Point Layer

Activation Function

relu

Embeds

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

illard

-0.85

redo

-0.76

linger

-0.75

jet

-0.70

burning

-0.69

ccoli

-0.69

lez

-0.68

ovie

-0.66

lishing

-0.66

paragraph

-0.66

POSITIVE LOGITS

Measure

1.00

 Measure

0.95

 measuring

0.89

 measurements

0.88

 measurement

0.85

 glim

0.83

 measure

0.82

 metrics

0.77

asured

0.76

 measures

0.75

Activations Density 0.026%

measurement-related terms

terms related to measurement and quantification

No Comments

No Known Activations